Muhammad Junaid Shah Bukhari

Machine learning engineer and NLP researcher with experience across industry internships and production grade ML systems. Specialized in transformer based NLP, computer vision, and predictive modeling with a strong focus on scalable and interpretable AI systems. Conducting independent research on 15 years of Pakistan National Assembly debates using BERTopic, Sentence Transformers, Word2Vec, and sentiment analysis for peer reviewed publication.

Focused on AI systems at the intersection of governance, policy analytics, and computational social science.

Machine Learning Engineer Intern - TechWave Analytics - Islamabad

(2025-06 - 2025-09)

• Developed supervised ML pipelines using Random Forest, XGBoost, and Gradient Boosting on 500K plus records, achieving up to 91% F1 score
• Reduced inference latency by 30% using SHAP based feature selection and optimized batch inference via Flask APIs
• Built reproducible data pipelines using Pandas and Scikit learn, reducing preprocessing effort by 60%
• Designed interactive dashboards using Plotly and Matplotlib for model evaluation and stakeholder reporting

Data Science Intern - Infotech Solutions - Islamabad

(2024-06 - 2024-08)

• Built NLP pipelines using TF IDF and Sentence Transformers for classification and clustering tasks
• Improved minority class recall by 18% using SMOTE and feature balancing strategies
• Performed statistical analysis including chi square, ANOVA, and KS tests for business decision support
• Integrated REST APIs with validated ETL pipelines for structured data ingestion

Bachelor of Science - Public Administration and Governance - NUML University